47 Years Later, The Palisades Disappeared Overnight
I grew up on Iliff Street, right in the middle of the ashes that up until a few nights ago, was a sunkissed neighborhood known as Pacific Palisades.
It was 1978, and I remember my dad climbing up on our roof with a garden hose. Every couple of hours, he would wet the house down, top-to-bottom, and everything surrounding it. I don’t remember everybody doing this, but my Dad is a Meteorologist, and back then he worked at the SCAQMD, the regional agency charged with studying, regulating, and improving air quality in Los Angeles, Orange, Riverside and San Bernardino Counties. Because of his specific remit and where we lived, he had a deep understanding of the Santa Ana winds and their effect on the Palisades.
When my dad explained what he was doing, he would point northeast to the hills behind us and tell us that if the winds didn’t die down, the fire miles in the distance would come towards our tiny little house and there would be trouble. As a small child, I don’t actually remember being scared about any of this. Every year there was a fire, the smoke was always so far away and so barely visible that it just seemed like anything else in life at the time. And besides, dads are superheroes to their children, so of course there was no danger.
We ended up living in the Palisades until I was 15, when we moved up to Seattle, but in that time, these sorts of fires happened almost every year to some degree. The 1978 one was a big one though, and my dad had a flight scheduled to New York the following morning. He woke up every hour during the night to check the wind readings, and loaded up our Impala wagon for a prompt evacuation. That is how precarious things were, even 47 years ago. By sheer luck that night, the winds subsided, and the Palisades were spared. A few more hours of wind and 1978 would have been 2025. Same damage. Completely flattened.
I guess Pacific Palisades has always been a wealthy area, but wealth back then seemed less of a step-function than it is now. The “alphabet streets” where we lived seemed like the most modest of the neighborhoods; the place where teachers, government workers, and clerks lived. My parents bought our 1200 square-foot rambler for $38,000, and pretty much no one in the neighborhood had views of anything. The neighborhood was so named because the streets went from Albright, to Bashford, to Carey, to Drummond, to Embury, to Fiske, to Galloway, to Hartzell, to Iliff, to Kagawa. Felt really great not having to even look that up just now, but I will admit that I was Today Years Old before realizing there is no J. I wonder why. Each street was basically the same, although I distinctly remember Galloway having more tree trunks pushing the sidewalks up, making it a better route for jumping your Mongoose bike or Powell-Peralta board.
As a kid growing up in the ’80s, the Palisades had everything you could possibly want. It was a “free-range” neighborhood, where we could ride our bikes as far away as the Santa Monica pier without worry. Palisades Park was home to AYSO soccer, tennis, basketball, “Par Course” stations, and most importantly to me, little league baseball. I was part of the “orange” franchise which started you off as a Ranger, then a Twin, and finally an Oriole. Our patron saint was Mel Haggai, who up until now, I didn’t know had flown 30 missions over Germany as a gunner when he was growing up (!!!). Mel got me to switch from first base to catcher, where I could have more impact on games. We didn’t have a single left-handed catcher’s mitt in the entire league, so he bought me one with his own money. I also remember perennial umpire John Meyers, who took me, my friend Adam Segel, and a few others to Disneyland, only to have to deal with us getting thrown out of the park for jumping off of a few of the rides and causing havoc.
We are Hiring at Microsoft AI
Since reflecting on one year at Microsoft almost 12 months ago, a lot has changed. Most visibly to the outside world, we’ve completely re-designed and re-engineered our Copilot consumer app from the ground up with a a craft- and quality-first mentality.
Every element moves more gracefully, every response loads much faster, and every detail is more considered. Copilot may be the tip of the spear when it comes to how we are thinking about our products over the next decade, but we’ve also begun to modernize Edge, Bing, and our news offerings. We are just getting started.
Here is a preview of a feature we are working on for Edge that allows you to co-browse the internet with your AI companion. It’s the most powerful addition to the web browser since video:
We aren’t just adding to Edge, but we are also — just as importantly — subtracting from it. Although Edge has grown browser share consistently for several years now, we’ve also grown our feature set a little too aggressively. Every so often, you have to step back from the bonsai tree, look at it from all sides, and prune. The Edge of tomorrow will be both simpler and more powerful than it is today.
Same deal over on Bing. We are pruning away old growth to make room for better-looking, more powerful search results:
There is a lot more going on in Microsoft AI as well, including a bunch of new things we are cooking up for MSN:
If you’re interested in joining the Microsoft AI team, we have a number of positions open right now that I’ve listed at the bottom of this post. If we already know each other, please reach out directly.
I mentioned in my last post a year ago how there were a lot of processes we needed to improve in the coming months. I detailed an unbroken chain of 8 steps for great products to emerge:
- An idea that can improve lives if executed well.
- Foundational research to light the path before design and coding begin.
- Rich design explorations and prototyping to make the experience palpable.
- Buy-in to build it at a level of quality that makes the team proud.
- Impeccable UX engineering and UX writing to make sure every detail is dialed.
- Well-conceived server-side engineering to make it scalable and maintainable.
- Creative marketing to prime people for the experience.
- … and finally, maybe more important that anything else on this list, the will to keep refining relentlessly after the experience is launched. This part is so often neglected as companies rush to build more things.
Although we are still in the first inning of this journey, I’m happy with the progress we’ve made across the board… in particular, steps 4 and 8. There is a different energy when everyone who is working on the product knows the standard is not “good”, not “better than our competitors”, but “the best version imaginable”. We are not there yet with any of our products, but there is no confusion where the bar is.
Reorienting our cross-functional teams to prioritize user experience has been a gratifying (and ongoing!) challenge. The vast majority of companies are not set up this way for a variety of reasons, some good, and some bad. In some industries, surviving might mostly about cost minimization, and not user experience. In other industries, “good enough” might be all people expect or need. But in just about all companies, there is some form of diminishing returns when it comes to how much time is spent getting every detail just right. The question is: where does a company draw the line?
One sign your company draws the line too low is if people insist that every detail, no matter how esthetic, be rigorously tested in order to prove it is instantly metrics-positive. This is a sign of a culture that rejects the value of taste and long-term vision.
Does that mean you should test nothing? Of course not. You should test everything, at the very least, to prevent accidentally messing something up. I still remember more than 20 years ago, during my first week at ESPN, causing our entire front page to disappear because I forgot a semicolon in a Javascript file. Testing for safety and information, though, is different than testing as religion. Furthermore, there are some things — particularly in the world of large language model training — that should absolutely follow a cycle of train/test/release/repeat.
Culturally speaking, this rebalancing of our values has created a spark for great work to emerge. Many of the same people who have been working on Bing, Edge, MSN, and Copilot for years simply have a new set of product-building principles now, in addition to some great new teammates and refreshingly smaller, evenly staffed cross-functional teams. It’s been wonderful seeing everyone exercise muscles that show up more often in startups and creative agencies than in 49-year old tech companies. The ingredients have been here all along, but the recipe is changing.
Speaking of ingredients, we are looking for more Designers and User Researchers to join the team at Microsoft AI.
These positions are United States only, Redmond-preferred, but we’ll also consider the Bay Area and other locations:
- Principal UX Researcher
- Design Director – News & Interests
- Principal Designer – News & Interests
- Senior Designer – News & Interests
- Principal Designer – Consumer Advertising
- Principal Designer – Growth
- Senior Designer – Growth
- Principal Designer – Rewards
- Principal Designer – Design Systems
These positions are specifically in our lovely Mountain View office:
Note that we are looking mainly for experienced, hands-on individual contributors to join our team. If inventing the design patterns for the next phase of the consumer internet sounds right up your alley, we’d love to talk to you.
We Live in the Golden Age of Ice Cream
I tried some absolutely outstanding ice cream yesterday that reminded me of yet another reason I feel lucky to be part of Generation X:
We are living in the golden age of ice cream.
In the 1970s, we had a few basic flavors to choose from: vanilla, chocolate, strawberry, chocolate chip, mint chocolate chip, and rocky road. There were a few shops like Baskin Robbins that marketed some scary stuff like Rum Raisin, but if you were just getting ice cream at the supermarket or some other ordinary location, those were your choices.
There also wasn’t as much science to ice cream back then. The term mouthfeel was still relatively obscure. It all seemed perfectly good back then because… ICE CREAM!!! … but compared to what we have today, it was, as the kids say, “supes basic”. In the 1980s, along came the frozen yogurt craze which pointed to how much untouched frontier was ahead of us. Shoutout to Humphrey Yogart, by the way — the world’s best named frozen dessert shop. Shoutout also to Aaron Cohen’s Gracie’s, which not only has excellent ice cream, but the world’s only bathroom dedicated to Dolly Parton.
Anyway, we’ve learned so much about fat content, texture, and flavor combinations since then that we have practically invented a new food group.
A few years ago, during a trip to the underrated gelato mecca that is Croatia, I had a variety which tasted different than anything I’d ever tried. It was a “fig yogurt gelato” from a little shop in Cavtat. I usually stay away from frozen yogurt because it’s generally not as good as ice cream, but this was intriguing. A great flavor (fig), combined with a beautiful tartness (yogurt), along with best texture for frozen desserts (gelato). It was outstanding, and I have not been able to find anything like it ever since…
Until yesterday!
Behold, Hellenika Cultured Gelato:
I don’t think I will ever purchase another brand of ice cream at the store ever again. It is the most perfect ice cream I have ever tasted. The creamiest mouthfeel, wonderful flavor combinations, and that little bit of tang which reminds you this is no ordinary substance.
Hellenika is a small creamery run by three Greek/Australian siblings with a single location in Pike Place Market in Seattle. Until recently, it was not available in stores, but you can now pick up pints in Metropolitan Market.
If you live in Seattle, you need to try this immediately.
If you don’t, you should try to get someone to mule you some in dry ice. Hopefully one day it will ship on Goldbelly.
Don’t get me wrong. I like a good low-grade Blue Bunny mini-cone as much as the next person, but I feel like we’ve passed through some sort of intergalactic hyperspace with this gelato. The future is now. Get this to your freezer any way you can. 🙌
One year at Microsoft
Last week marked one year at Microsoft for me, and what an unexpected adventure it’s been! I thought I was coming in to lead a a stable of popular, but well-trodden web properties, and I ended up getting to work on a whole lot more, including Windows, Bing Chat, and the company’s biggest bet in years: Copilot.
I usually write a lot about the companies I work at but have held off until now because we haven’t been hiring. Well now we are! We’re specifically looking for Designers and UX Engineers to work on our design system for Copilot. Some of these positions are on my team and based where we have offices (Puget Sound, the Bay Area, Atlanta, New York, Vancouver, Barcelona, Hyderabad, Beijing, and Suzhou) and some are on adjacent teams and can accommodate fully remote work. If you are a Designer or UX Engineer with a passion for design systems and AI, we’d love to chat.
So what has year one been like? The good and the “needs improvement”. 🙂 👇
One of the reasons I decided to join Microsoft was I missed the joy of in-person product-making. I know not everyone feels the same way so I’m not trying to make any broad statements about local vs. remote work, but for me, it has been even more refreshing than I expected. I usually come in 3-4 days a week, while others on the team are anywhere from 0 to 5.
It’s funny, sometimes I will wake up on a Monday and think to myself “ahhhh, this is going to be a chill work-from-home day” and by the end of the day, I realize I’ve been staring into a screen on video calls for almost the entire day and how much that slowly saps my energy. Meanwhile, in-person days are filled with walks, whiteboarding, and energizing sessions with some of my favorite teammates I’ve ever had. I realize not everyone feels this way about being in the office from time-to-time, but I do. Even our fierce, interdepartmental karaoke battle helped bring a bunch of teams together who had never met before.
The other great thing about lucking out and joining when I did is that we are embarking on one of the rare paradigm shifts that occurs in technology maybe once a decade. The 1980s were about personal computers. The 1990s were about the internet. The 2000s were about smartphones. The 2010s were about the cloud. And the 2020s will be about AI. The really powerful thing about all of these developments is that they don’t replace each other, but rather they build on each other. AI is the result of everything that came before it. If you got into design to help shape the culture of the world around you, these are the moments you treasure.
These turning points are also wonderful because they give you a chance to reawaken to your Beginner’s Mind. I have learned more in my first year here than at almost any other time in my career. Not only does the pace of technological change in AI force you to build skills as you go, but there are so many amazing engineers, designers, researchers, writers, marketers and other creative people to learn from, that it happens almost automatically.
When I was interviewing here, not even my prospective new boss told me about any of what was behind the curtain. Only during my first week did I find out all we are working on to empower every person and every organization on the planet to achieve more. That is Microsoft’s mission statement, if you hadn’t heard it before. It’s an uncommonly good filter with which we can all ask ourselves every day “does this project actually do that?” It’s quite freeing as it gives you license to question projects at every stage of development.
Another thing I’ve loved about my first year here is that I joined a group that has figured out how to ship very quickly. That also has its downsides, as we have plenty of craft problems to solve, but it’s great working for one of the largest and most established companies in tech and being able to ship within weeks of designing something.
One more thing that’s blown me away is the Inclusive Tech Lab, where we work on new technologies to make our products inclusive to people all of abilities and walks of life. No one experiences technology the same way, and teammates like Dave Dame and Bryce Johnson do a ton of great work to make sure that’s top of mind for everyone.
Finally, one of the unsung benefits of working for a native Seattle company again is that Seahawks stuff is all over the place. Presentations, charity auctions, everyday office attire… you name it. It’s nice to not be the only one with good taste in football teams.
I am no corporate shill, however, so I must also be honest about some of the things that need improvement over here.
At the top of my list is that Microsoft has not yet fully embraced the role design plays at most other tech companies. We were engineering-driven in 1975 and we are squarely engineering-driven in 2023. The world, meanwhile, has changed in that time. It is no longer sufficient for complex things to work. They must also shed their complexity. People expect the products and services they spend their time and money on to delight. To overdeliver. To give them superpowers. Those sorts of qualities only materialize when you have supergroups building products.
In music, a supergroup is when a singer, a guitarist, a bassist, a drummer, a keyboardist (and so on) who are all at the top of their game come together to create an album. In tech, a supergroup is a researcher, a couple of designers, a product manager, and a Volkswagen Bus or two full of engineers. Up until a decade or so ago, a lot of tech companies followed the model of packing projects with as many smart engineers as they could find and only sprinkled in things like design and research as necessary. I still see some of this thinking in pockets over here. I’m trying to influence things, but it’s a delicate dance, especially when the company has had such enormous success by doing so many other things very well.
I think there are plenty of people here who still feel like being engineering-led is unequivocally good, but to those people I would say that in a modern tech company, design is engineering. It’s no better or worse, but it does have very different leverage in the building of a product.
On the plus side, Microsoft has never had as much design and research talent as it has right now, and we are increasingly looked to by the executive leadership team as lighters of the path. When we go into high-stakes meetings, we always go in with pixels and prototypes, which are uniquely good at cutting through bullshit and ambiguity. As a wise person once said, a prototype is worth a thousand meetings.
There are a ton of amazing designers and researchers who have been here for 10 and even 20+ years whose hard work has led to this moment of evolution for the company, and every day I am in awe of their perseverance.
Next on my list is how often we get in our own way with “procedural goo”. I’ve worked in plenty of large companies including Twitter, Disney, and MSNBC, and have never seen the level of approvals, paperwork, and rules that get in the way of speed and autonomy here. Just transferring someone from within my own org to a slightly different role within my own org took a dizzying amount of effort. At most companies, this would have been about 10 minutes of work: one minute from me and nine from someone in HR trying to navigate to the right screen in Workday.
Then we have acronyms. My GOD do we have acronyms. I actually liked acronyms before I got here! I usually think they are cute. After seeing a new one almost every single day since getting here, I have resolved to never use them either inside or outside of work. I even say “Cyan Magenta Yellow Black” out loud if I have to!
Finally, the last thing on my list — and this is where you come in — is dedication to craft. It is so tempting to try and “science” your way into viable products these days. Build the beginnings of a customer base through rudimentary product-market fit, and then fastidiously optimize your funnel, your game mechanics, your viral loops, your push notifications, and so on and so forth. These gains are not always easy to come by, but they are rooted in ruthless experimentation and allegiance to short-term data. Our north star is at least pretty pure — Daily Active Users — and that metric is usually a good indicator that you’ve made something people like, but doctrinaire allegiance to almost any singular metric can quickly make people forget why we are in this profession to begin with: to improve lives. Or to put it squarely in Microsoft parlance again: to help every person and organization on the planet achieve more.
If you ever find yourself asking the question “how can we increase Daily Active Users?” instead of “how can we make our product better for people?”, you’ve already lost. Metrics are trailing indicators of qualitative improvements or degradations you’ve made for your customers… they are not the point of the work.
Recently, we’ve made some excellent strides in prioritizing qualitative product improvements even when they fly in the face of metrics we care about, and it’s really gratifying to see. It reminds everyone that a product is the collision of thousands of details, and the crafting of these details requires taste.
Designers are often looked to as “owners of craft and taste”, but craft is very much a team sport. It’s not just how things look and feel but also how they work. I very much like how Nick Jones (channeling Patrick Collison) at Stripe put it in this video:
“What we put out there should quite plausibly be the best version of that thing on the internet“.
To do this takes an unbroken chain of excellence:
- An idea that can improve lives if executed well.
- Foundational research to light the path before design and coding begin.
- Rich design explorations and prototyping to make the experience palpable.
- Buy-in to build it at a level of quality that makes the team proud.
- Impeccable UX engineering and UX writing to make sure every detail is dialed.
- Well-conceived server-side engineering to make it scalable and maintainable.
- Creative marketing to prime people for the experience.
- … and finally, maybe more important that anything else on this list, the will to keep refining relentlessly after the experience is launched. This part is so often neglected as companies rush to build more things.
Some people would look at this list and think “yep, makes sense”. Others would look at it and think “sounds slow and not very agile”. The trick to balancing this level of quality with speed of development is realizing that it’s often more efficient to experiment in step 3 than it is in step 6. This is why so many modern tech companies realize that hiring more designers and researchers doesn’t waste time and money… it saves time and money. With less than a week of one designer’s time, we can produce a wide variety of prototypes to test with real people. Just recently, we created an entire GPT-powered research application without even bothering a single engineer.
Design is engineering.
Finally, a brief note about prototyping. I would argue that the most impactful innovation in the craft of product development over the last 20 years has been the rise of rapid design prototyping. Prototypes that demonstrate an experience are useful not just in usability testing, but also in selling ideas up and across the organization. Engineers hate working on things that haven’t been thought through or “appropriately politicked” yet, and if you can bring them a working prototype that has already been vetted with users and various stakeholders across the company, they will love you for it and work hand-in-hand with you to get every detail right.
Design prototypes are the currency of a high-craft, high-speed product development organization, and they are increasingly the currency of our team.
Alright, back to the hiring. I plan to hire against the entire growing list of products our team is responsible for: Copilot, Windows, Edge, Bing, Start, Skype, SwiftKey, and so much more… but for now, this is a concerted hiring effort centered around Designers and UX Engineers to help build out the emerging design system for Copilot and our suite of AI-powered products.
If this is you, please have a look at the following roles we’ve just posted:
- Designer II
- Senior Designer
- Principal Designer
- UX Engineer II
- Senior UX Engineer
- Principal UX Engineer
We’d love to work with you on the future of design systems at Microsoft!